Richards, J.E. (2018). Development of diffuse optical tomography sensitivity in infants. International Conference on Infant Studies, Philadelphia, July, 2018.(pdf )
Development of diffuse optical tomography sensitivity in infants. ‘Near-Infrared-Optical-Spectroscopy’ (NIRS) is a tool for neuroimaging in infant participants. NIRS measurement works by emitter/detector optodes placed on the scalp that measure reflected light from oxygenated / deoxygenated hemoglobin. Diffuse optical tomography (DOT) describes the scattering of light through the interior of the head. The sensitivity profile derived from DOT may be used to identify the underlying cortical anatomy that is reflected to the detector optodes. The current study used simulated photon migration methods to map the DOT sensitivity of infants in the first year. The DOT sensitivity profiles were used to complement spatial projection methods for a comprehensive database of scalp-location-to-cortical-anatomy for infants. The DOT sensitivity volumes may be used for inverse modeling of cortical activity from NIRS recordings (e.g., Homer2 and AtlasViewer) for either individuals or average MRI templates from the Neurodevelopmental MRI Database. The structural MRI volumes came from individual infants and average templates (3, 4.5, 6, 7.5, 9, and 12 months; also 2, 4, 12, and 20-24 years) from the “Neurodevelopmental MRI Database”. The DOT sensitivity was estimated on each MRI with photon migration simulation programs (e.g. MCX; tMCimg; MMC) by projecting 100M photons from scalp locations into a segmented MRI volume and recording the flux from the projection at each voxel in the MRI. The scalp locations were scattered semi-uniformly on the scalp (10-5 EEG placements). These locations were used as emitter/detector DOT channels centered on the 81 locations of the 10-10 EEG placement system. Figure 1 shows the fluence (summed flux) as a function of the distance from the 10-10 locations for a 6-month-old participant and for the average over all participants for this age. The peak, slope derivative location, and half-height-half-width location are marked on the graphs. These results imply that about 15-22 mm is the maximum depth sensitivity for NIRS recording (bottom panel); though this differs across the head (top panel). The fluence level was smaller and the sensitivity profile was flatter for emitter/detector distances from 30 to 40 mm, than from 15 to 29 mm. Figure 2 shows the avg peak depth for the MCX DOT sensitivity profile over age. There was a decrease in the depth of the peak location from 3 to 12 months, and this increased from 2 years through adults. The DOT sensitivity profiles were used to estimate the contribution of cortical ROIs to the NIRS signal recorded on the scalp. Stereotaxic atlas volumes were used to define the cortical ROIs. Several methods for detecting the ROI contribution were compared (location of max DOT. average DOT in 3 mm sphere, distribution of DOT sensitivity across ROIs). ‘Lookup tables’ were constructed that have anatomical locations from three stereotaxic atlases separated by each 10-10 location, age, average or aggregated individuals. There was good consistency for the ROis across the photon simulation methods, and DOT sensitivity locations. The DOT sensitivity profiles provided a quantitative measure of the photon sensitivity across the multiple ROIs for a single emitter/detector channel.